Feeder controller board redesign

[trev]

The solenoid seems to move pretty quick, maybe 10s of ms. I read up a bit and you need that kick to get the thing moving. Once it has moved into position you can reduce the current/voltage to hold it in position but the initial movement needs to be good. I do not know how much variation there is between a cold day and a warm day. Or even from solenoid to solenoid production variation.

Maybe the original pins do not need to be machined if the voltage level is set correctly. I am hoping, once the PCBs arrive I can build one up with the secondaries connected in parallel. I think it really only needs one secondary connected now as the current consumption is greatly reduced. Probably I can wire across the one bridge rectifier and test it out. Might be as simple as that. But if the pins do need to be changed that is not the end of the world. It would certainly make replacing the controller board a lot easier if the pins did not need changing. 

But if I also need to use a bench supply and program the voltage to just be on the edge each time, that is not the end of the world for me. I mean 18V-20V at 400mA is nothing really. All the feeders could be powered from a basic bench supply very easily. The transformers could be completely removed. The big cap on the board would help with solenoid current draw but everything else I reckon would be pretty stable. I do not think there would be much voltage drop in the cables powering the feeders from a supply. 

This would be totally over the top but be able to serve all the feeders connected and is pretty cheap really.

https://cpc.farnell.com/tenma/72-10480/power-supply-1ch-30v-3a-adjustable/dp/IN06822?st=bench%20power%20supply

Trev

[trev]

Not sure how long it would last but this is dirt cheap

https://www.amazon.co.uk/SHNITPWR-Universal-Power-Adapter-Transformer/dp/B091L5GB6W/ref=sr_1_6?dchild=1&keywords=adjustable+power+supply&qid=1631529962&s=electronics&sr=1-6

[trev]

This panel mount one looks interesting. It could be installed inside the RV power enclosure and then feed the cable so no external cable alterations would be required. Then you can adjust the voltage from the panel. Nice and tidy.

https://www.amazon.co.uk/Voltage-Regulator-Programmable-Adjustable-Converter/dp/B08P7PG7R5/ref=sr_1_37?dchild=1&keywords=adjustable+power+supply&qid=1631530107&s=electronics&sr=1-37

Anyone know if these blow up a lot?

[Mike]

From memory ISTR the transformer has 2 windings in series - could it be as simple as just using one winding to reduce the voltage? I would think the stepper would be OK with the reduced voltage with speed ramping and the smoother control the Trinamic driver gives.

[trev]

I have now done around 2000 placements with a feeder installed with the new controller design. It has not missed a beat and appears to be performing very very nicely. Lovely smooth movement means the components are always in the same position and I think this helps placement be that little bit more repeatable. Also being so quiet is fantastic.

I was not using lots of plastic tape due to the job I had to get done. It was mainly paper but there were 2 plastic tapes on the bank. They were quite thick, the one held a SOT23-5 and the other a 1210 inductor. The SOT23-5 moved very smoothly with not shaking in the tape. Position again was very repeatable.

What I did find out though is the reduced voltage for solenoid operation is because I did not fit the extra springs on the grey feeder. This style feeder has 2 sets of springs to help stop the pins sticking and I did experience a bit of pin sticking. When I checked a different feeder I found the solenoid voltage needed to be up at around 24V to work reliably due to the extra force of the springs. Still seems that 37-39V is quite high.

The other thing is that with higher voltage the solenoid does not like to be energised for long. With the original design moving the tape quickly, the solenoid was on for less time and this may have helped keep heat down. We this design design if the voltage is less this might balance out.

I have another mod to try to stop the pin hitting the plate that does not require any machining of the pins. If this works then I think I will have a final solution.

Finally I have attached a pic of the first hand soldered controller board built from PCBs I got made. Everything appears to be working as expected.

[trev]

I did a run of a few boards today and compared using a feeder with an old control board in it with one with my new design. I have to say, the old feeder is so fast. I am not sure why they felt the need to move the tape so quickly though. They have so much time between placements! Maybe when using merge moves or something and the feeder is right next to the camera, it needs to be. The speed can not help when winding the tape back though. I much prefer my new design.

I am going to build a few more boards up and convert some more feeders. I then have a big job to get done and this should really put the board through final testing.

Trev

[Mike]

I am not sure why they felt the need to move the tape so quickly though.

For multiple feeds perhaps.

[trev]

Been a while since I did work on this. All my attention up until today had been with a grey feeder where the positioning pin is right at the front of the feeder. I was still finding that despite motion being good everything else was rubbish. I just could not get reliable movement and was feeling beaten. The tape would lift up sometimes and the pin not move it. With too little tension on the pin the tap would still bounce about. I was a bit gutted after all the work I had done. I felt it was never going to work. It was working perfectly for SOT23 and thicker tape but 0603 LEDs was still bad.

Today though I found the motivation to try again but with a blue feeder. These feeders have the positioning pin at the middle of the feeder right next to the rubber band. This seems to have made a massive difference. With the super smooth motion my board created there is no vibration of leds when the tape is moved. The stepper motor gearing does not chatter at all. But what's different is the positioning pin  is so far from the open end of the tape that it does not affect the leds at all when it snaps up and down. So the original issue with blue feeders I think was mainly the movement rather than the action of the pin going up and down. With the grey feeders the issue is that movement and also the pin being so close to the open end of the tape that it caused vibration when snapping up and down.

I placed 350 leds today and had no pickup failures. It was perfect. Will keep using it but it is looking very positive finally.

Have attached link to led movement in action. It is super smooth every time which I also think helps reduce band slippage and also reduces the risk of the pin stretching the hole in the plastic tape and causing movement failure. 

https://youtu.be/0Wu516kW94U

Trev

[spiyda]

https://youtu.be/0Wu516kW94U

Trev

Looking very nice

[trev]

Side view of a feeder in action using my new board. It also shows the mechanism for anyone who hasn't opened a feeder.

https://youtu.be/JyTG3TL-eME

[trev]

Used the controller board with several feeders and reliability of the board is excellent. Specifically regarding 0603 leds which was the main reason I designed a replacement controller, have placed 1400 leds and the movement was perfect everything single time.

[Greasemonkey]

Hello Trev,
I was wondering what happened to your controller project.
I am interested in changing the drive boards in our old workhorse's feeders.
It would be good to have a drive board that does ramping or smooth as we do find certain components can jump around.
Did you have any plan to make these boards available to others or was it just a pet project?

[trev]

Hi

I used the board on several feeders and it performs fantastically well. I think I placed 1000s of 0603 leds and jumping was greatly reduced. Some feeders have the pins at the front ( grey feeders ), some have them at the back ( blue feeders ). I found on the grey feeder some times the action of the pin would also contribute to the bouncing. So if you have blue feeders then I think it works better. It might be other things too but I think that's what I found.

The solution I used as I was investigating the problem and using the boards was to wire the transformer inside to power the control board but then I used an external supply to power the solenoids. This allowed me to monitor them and see what voltage worked best. Reducing the voltage to experiment with the pin impact. I think though in the end I settled on 24V and it was find for a lot. With the transformer, the voltage is all over the place and I think too high most of the time.

Also I found that when you start to slow the feed speed down, the solenoids are energised for longer and the transformer voltage is too high to turn the solenoid on for that long. 24V enables you to slow things down and not overheat the solenoids.

The software for the controller is Arduino based and had a menu to control everything. I ended up putting a panel mount usb connector on the back of the feeder and I would plug into that to configure it from a terminal programme.

It was always a work in progress but was at a stage where it was working. I do have some boards and would be happy to provide all the source code and schematics so you could modify/adapt it to your own needs.

How many boards would you be looking for? 

[Greasemonkey]

Hi Trev,

Thats a good question.
I am thinking initialy I am after a few boards to try out and tinker with.
If this works out well then I would probably want to retrofit most of our feeders down the line.
I have a cache of chinese arduino megas collecting dust in my stores from old venture which would be ideal.

We still use our RV a lot for certain jobs and I am on a bit of a mission to fix or improve a number of the known issues we suffer with the feeders which prevent an operator walking away from the machine while it is running.

It would also be nice to add some features to the feeders which would make life easier.

[trev]

I have a few feeders that are already wired up and some spare boards. Might be useful having some already fitted for reference and to enable you to experiment before taking apart your feeders? And to have some more feeders.

I would likely sell these relatively cheap if you are interested.

Do you need any other spares whilst at it as I am having a final clear out and may put some stuff on eBay in the new year if I can't sell them to someone on here for a good price.